Re: [patch] Race between init_idle and reschedule_idle

[Richard Gooch]

Martin J. Bligh writes:
> Thanks to Alan Cox & Andrew Morton for showing me how to serialise
> the cpus to make the panic legible. The following patch holds back
> the boot cpu at the end of smp_init until all the secondarys have
> done init_idle:

One thing that bothers me about your patch is how it limits the number
of CPU's to the number of bits in an unsigned long. While I realise
there are other places that do the same (last time I looked), we
shouldn't be perpeturating these kinds of limitations.

I'd suggest you use an atomic_t instead. Increment for each CPU, and
decrement when each CPU is ready. Just test for 0 in your wait loop.
One less piece of code we have to overhaul later.

				Regards,

					Richard....
Permanent: [e-mail]
Current:   [e-mail]